Browsing by Author "Aho, Varpu"

Mastitis is economically the most important disease and the second most important welfare issue after lameness in dairy production worldwide. Mastitis diagnosis consists of recognizing the causative pathogen and simultaneous changes in milk parameters, such as somatic cell count. Currently, 27 % of Finnish farms use automatic milking system (AMS) and more than 50 % of all milk is harvested by a milking robot. Large amounts of data are available from AMS, and they can be used to recognize and control mastitis on farms. The aim of this work was to study how different AMS data patterns describe mammary gland infection, and how they can be used in mastitis diagnosis. The most conventional parameter for diagnosing mastitis is somatic cell count (SCC) which describes the number of somatic cells per milliliter of milk. During mastitis, SCC increases, but a significant day-to-day variation is characteristic. SCC is measured in official Dairy Herd Improvement (DHI) programs, and SCC is also counted by sensors in AMS. The most common in-line measured parameter at AMS is electrical conductivity (EC). EC is measured quarter-specifically which makes it good for comparison among different quarters but there are some uncertainties associated with EC. In addition, L-lactate dehydrogenase (LDH) is an enzyme that indicates infection in different tissues and is also detectable with a sensor in some AMS. It’s less mastitis-specific than SCC, but because it has less daily variation, combined with SCC it’s currently an interesting tool for recognizing mastitis in AMS. Descriptive study was conducted using AMS data from 24 cows over 7 months from a Canadian research herd. The data were fragmented and only a few mastitis cases were included. However, the results describe the characteristics of different AMS parameters. Results showed that LDH is high especially in 1st lactation cows until 35 days after calving. As expected, LDH of mastitic cows was substantially higher compared to cows that were healthy or had non-udder illness. Interestingly, the daily variation of LDH in individual cows appears to be greater than expected.